ABSTRACT
Introduction This paper discusses the usefulness of currently available information on the poor, for designing public action against poverty in Africa and the Middle East. This issue is raised against the background of recent arguments that emphasize economic growth as a means of alleviating poverty. There is little debate that a link between growth and poverty exists. It is to be emphasized however that growth is not all that matters for poverty reduction. Firstly, even if economic growth rates explain 50% of the variation in poverty incidences as some recent research suggests, 50% is still unexplained, indicating a large margin for poverty-reducing policy. Secondly, studies reveal that the growth-poverty elasticity varies across regions and is the lowest in Africa. Thirdly, the “poor” cannot be regarded as a homogeneous group. In Kenya and Nigeria for example, extreme poverty increased while “moderate” poverty decreased, suggesting that the poorest profited less from economic growth. Overall in Africa, as to be discussed later, the poverty gap (average shortfall below the poverty line [PL] of the poor) worsened faster than the headcount index (share of population below the PL). This indicates that not only did more people slip below the PL, but that even among those already poor, many got even poorer. In the Ivory Coast for example, even in the context of a rapidly contracting economy, many of the poor were no longer poor a year later. 1 Therefore, anti-poverty policies have a role to play. For anti-poverty interventions, policy-makers and program designers require accurate information on the poor. Without accurate and timely information, it is difficult to know whom to target with anti-poverty measures, and know who benefits from specific policies. Different antipoverty interventions involve different levels of targeting. This results in
different demands for information on the poor. Sectoral targeting, i.e. spending in sectors which are relatively more important to the poor (e.g. primary health care), is the least demanding in terms of specific information about the poor. Self-targeting, by definition, does not need such detailed information. Nevertheless, key information is still essential to know where to implement these programs for example, and to properly evaluate them. Several anti-poverty policies rely on administered targeting, which requires detailed information on the poor, to determine indicators or socio-economic categories highly correlated with poverty. Often, the record of administered targeting has not been good. In Burkina Faso during the mid-1980s, famine early warning systems were used to target food aid to the arid northern Sahelian zone. Yet, Sahelian households had higher and more diversified incomes than households in the more favorable agro-ecological zones in the south. As a result, better-off households living in the Sahel, received ten times more food aid than more vulnerable households in the south (Reardon et al. cited by Lundberg and Diskin 1995: 16). Examples like this illustrate that fairly detailed knowledge is essential for avoiding targeting errors, such as leakage to the non-poor and imperfect coverage of the poor.This paper discusses the information commonly available on poverty in Africa and the Middle East. Conclusions about trends in poverty and socio-economic profiles of the poor depend on the definition of poverty used. Some definitions have centered around broader notions of human capabilities combining health, education and income indicators (e.g. UNDP’s Human Development Index). Other definitions have been narrower, focussing on single yardsticks of welfare. Other approaches have attempted to leave definitions of welfare to local communities themselves, rather than adopt externally imposed criteria. Each definition presents a degree of uncertainty whether some “justifiably deprived” people may have been ignored. Quite often, differences emerge in the groups identified as poor, and sometimes the differences are irreconcilable. Also, poverty definitions may be used to alter purposefully the recognized constituency of the poor (e.g. Ukraine Human Development Report 1996: 28).The importance of definitional choice in poverty analysis is not underestimated which is later discussed. However, the focus is on inadequate consumption as a key element of poverty, supported by a few other indicators. In choosing other indicators, “output indicators” (e.g. life expectancy) are highlighted rather than what may be regarded as “input indicators” (e.g. access to health care). Such data are even less available than data on consumption-poverty. For practical reasons, the authors had to rely on international sources for information. This means that this report on
poverty information is not as comprehensive as it should be. It is assumed however, that international efforts to gather information, specifically in Africa, would have uncovered significant sources of poverty information of high quality.
